Moving in the other direction? The impact of domestic party system change on Italian MEPs

Transcription

1 Journal of European Public Policy 11:6 December 2004: Moving in the other direction? The impact of domestic party system change on Italian MEPs Amie Kreppel ABSTRACT This paper re-examines some of the assumptions of the Europeanization literature by suggesting that it might be possible for domestic political events to have a direct influence on supranational political organizations. This possibility is addressed through an analysis of the impact that the dramatic political reform of the Italian party system in the 1990s had on the primary home of political party action in the EU, the European Parliament. The goal is to determine the extent to which the transformation of the Italian political party system impacted the party system of the EP, including internal party group cohesion and the political roles and influence of Italian MEPs within the EP. The results are in many ways counterintuitive and suggest that the norms and rules that structure the supranational institutions may be hardier and more resistant to shock than currently assumed. KEY WORDS European Parliament; Europeanization; Italy; party system reform. There can be little doubt that the development of the European Union (EU) has had a tremendous impact on the domestic political life of its member states. The flourishing field of Europeanization literature clearly examines and demonstrates this impact in a wide range of spheres. This paper seeks to ask a different question, or rather to reverse the question to ask whether, or to what extent, political changes in the national sphere affect the European political arena. Although some aspects of this relationship are quite clear, others are less obvious. This paper seeks to examine this question through an analysis of the behaviour of Italian Members of the European Parliament (MEPs) before and after the electoral reforms and political upheavals of the 1990s. The dramatic transformation of the Italian political scene and the political party system unquestionably reshaped the national political landscape, but did these changes have any impact at the supranational level? This paper traces the shifts in the behaviour and roles of Italian MEPs within the European Parliament (EP) before and after the transformation of the party system during the 1990s, Journal of European Public Policy ISSN print; online 2004 Taylor & Francis Ltd DOI: /

2 976 Journal of European Public Policy taking as its starting point the assumption that national parties, as sole gatekeepers of the EP electoral lists, have the potential to significantly impact MEP behaviour (Hix 2002; Kreppel 2002b). While the immediate subject of the paper is the potential changes in the behaviour and roles of Italian MEPs and the influence of domestic political reforms on supranational actors, the broader subject addressed is the extent to which Europeanization is a one-way street or a recursive relationship between Europe and the member states. This question takes on increased relevance in light of the enlargement of the EU with ten new member states. The domestic politics of the newly democratic countries of Eastern and Central Europe are markedly different from those of the current member states. Party systems in particular are more fluid and parties play a substantially reduced role in most cases. Thus, the impact of party system change and large-scale political reform in Italy on Italian MEPs may be understood in the broader context of EU enlargement as an example of potential future patterns among MEPs from the new member countries. In this paper I first examine why we might expect national political events to impact the supranational level. I explain why Europeanization might move in the other direction through an analysis of national political party competition and its potential to impact the EP. In the second section I briefly summarize the political upheaval experienced in Italy from the late 1980s through the mid-1990s. The following sections outline in detail the specific questions addressed by the data analysis, the data itself and the process of analysis, while sections four and five report the results and interpret their meaning. In the final concluding section I focus on the extent to which the lessons of this specific research can be broadly applied or generalized and what this might mean in terms of EU enlargement. MOVING IN THE OTHER DIRECTION? The focus in recent years on the impact of European integration on the politics, policies and institutions of the member states of the EU is understandable given the rapid increase in the overall pace of integration since the late 1980s. Literature on Europeanization has emphasized, among other things, the impact of EU integration on specific national policy arenas (Radaelli 1997; Ágh 1999; Knill and Lehmkuhl 2002; Harcourt 2002), institutions (Ladrech 2001; Cole and Drake 2000; Grabbe 2001; Hanley 2002), regions and regionalism (Börzel 1999, 2000, 2002; Benz and Eberlein 1999) and even political systems as a whole (Ladrech 1994; Anderson 2002). Almost universally, however, the focus has been on the impact of the development of EU-level policies, institutions and overall integration on the political systems of the member states. The causal arrow has consistently pointed from the supranational level down toward the member states, despite frequent recognition of the fact that national actors often use the supranational institutions as tools through which they can

3 A. Kreppel: Moving in the right direction? 977 indirectly influence domestic politics (Hix and Goetz 2000; Ladrech 2002; Cole 2001; Anderson 2002). In fact, many definitions of Europeanization themselves suggest at least the possibility of a more recursive relationship between European (EU) and national level political development. Indeed, the notion that domestic politics may impact European level decision-making and outcomes is not new. This idea is an integral part of the intergovernmentalist theoretical framework, which posits national leaders and interests as the primary forces within the integration process (Moravcsik 1993, 1998). The intergovernmental approach highlights the possibility for changes in national leadership to have significant short- and long-term effects on the process and content of EU integration. 1 One need only think about such crucial actors as Charles de Gaulle, Konrad Adenauer, Margaret Thatcher, François Mitterrand and Helmut Kohl to realize the positive and negative effects that individual national leaders can have and have had on EU integration as a whole. But while few would dispute that national leadership matters for the future path of the EU, the potential impact of substantial national party system change remains unclear and largely unstudied. This is surprising given the recent deluge of literature on the growing importance of the EP and the critical role of parties within it. Although some disputes remain over the particulars, there is now broad agreement that: (a) the powers of the EP have grown substantially since the Single European Act (SEA), (b) the EP today merits the title of co-legislator and acts as a functioning legislative body in most significant policy arenas within the EU, and (c) the political party groups are the primary organizational unit within the EP (Tsebelis 1995; Kreppel 2000, 2002a; Hix 1999; Corbett et al. 2003). If the EP is a crucial aspect of EU decision-making and the political party groups are the dominant force within it, why shouldn t we be concerned about the development of those party groups and in particular the impact that national party system change might have on them? To understand the direct relationship between national party systems and the party system of the EP it is important to remember that EP party groups are fundamentally conglomerations of national level parties. The supranational party groups do not exist at the national level. Elections to the EP are based on nationally derived electoral lists. MEPs once elected on the basis of their national party decide (generally within their delegation, although individuals may act on their own) which EP party group to join. Alternatively, national delegations may decide to remain unaffiliated or even work to create a new party group. Even after they have joined a party group the national delegations may influence the activities of the EP as a whole, and will certainly exert at least some influence over their own EP party group (Kreppel 2002b). It has even been argued that, amongst the three possible sources for influence and control over MEP voting behaviour, personal preferences, EP party group preferences and national party preferences, the latter takes precedence and best explains variation in voting patterns (Hix 2002).

4 978 Journal of European Public Policy As a result significant changes in domestic party systems could potentially influence EU outcomes (policy and other) through the activities of the national party delegations within the EP and at the very least should impact the behaviour of the affected national delegations within their respective party groups. The impact of national change is more likely to be significant in those countries that have substantial representation in the EP given that many crucial activities require high vote thresholds and the general application of the proportionality rule means that size matters. 2 Italy, as one of the current big states had eighty-seven members in the EP representing 14 per cent of the EP during the period analysed here. 3 It seems worth investigating, therefore, whether the dramatic reforms and party system upheavals in the 1980s and 1990s in Italy had any impact on the EP and the supranational party system through significant changes in affected Italian MEP behaviour and likewise whether these changes had any subsequent impact on the ability of Italian MEPs to wield influence within the EP. 4 Before moving on to examine this influence directly it will be useful to briefly review some of the changes in the Italian political system over the last fifteen years, focusing in particular on the transformation of the party system. THE ITALIAN PARTY SYSTEM IN UPHEAVAL The depth of the political upheaval that took place in Italy between 1988 and 1995 has led some to call the outcome the Italian Second Republic (D Alimonte and Nelkin 1997; D Alimonte and Bartolini 1995, 2002; Bufacchi and Burgess 1997). A full analysis of the transition is beyond the scope of this paper. Instead, this section briefly reviews four aspects of party system change that occurred during this period. Ω The voluntary dismantling of the Communist Party (Partito Communista, PCI) and creation of the Democratic Party of the Left (Partito Democratica di Sinistra, PDS). Ω The implosion of the Christian Democrats (Democrazia Cristiana, DC) with the subsequent birth of a number of small parties of the centre formed largely around personalities. Ω The rise of the centre-right Forza Italia party. Ω The wholesale restructuring of the party system in terms of inter-party relationships as a result of the reform of the electoral system. The first signs of change began in 1988 with the initial efforts to reform the PCI in the aftermath of the Tiananmen Square tragedy; however, it was not until the fall of the Berlin Wall and the collapse of the Soviet bloc that real change occurred. The process began with a series of reforms passed by the PCI central committee in November These included a new name the Democratic Party of the Left (PDS); a new symbol the oak tree; and a significant political realignment towards the centre-left. The implementation of the proposed reforms was not completed until 1991 and failed to lead to

5 A. Kreppel: Moving in the right direction? 979 much support among voters due to the rise of factionalism within the party, the slow implementation of reforms, and the absence of a clear party agenda. In the general elections of April 1992 the new PDS received 16.1 per cent of the vote, down considerably from the previous elections in 1987 when the PCI received 26.6 per cent. 5 The transformation of the Christian Democrats (DC) also began in 1989 as its anti-communist stance, a key aspect of its platform, became increasingly obsolete. Combined with growing allegations of corruption, failure to reform the political system and the inability to counteract Italy s economic crisis, this weakness began to threaten the party s future. The 1992 general elections gave a clear indication of the trouble that the government parties, and especially the DC, were in. The outgoing four-party coalition led by the DC barely retained a majority of Senate and Chamber seats but its combined share of the vote fell below the 50 per cent threshold and for the first time in its history the DC did not receive at least 30 per cent of the vote. The widespread delegitimization of the DC due to the expanding corruption scandals led to attempts to stabilize and rehabilitate the party by renaming it the Italian Popular Party (PPI) and redefining its goals and internal organization. However, all failed to save the DC and by early 1994 it had split into four parties: (1) the PPI, (2) the Christian Democratic Centre (CCD), a more conservative party, (3) the Pact for Italy (PI), based on liberal democratic reform, and (4) the Social Christians (C-S), a centre-left oriented group. 6 The implosion of the DC continued in 1995 when the still new PPI split in two, followed closely by the creation of the Rinovamento Italiano (a centrist party) and the creation in early 1998 of the Unione Democratica per la Repubblica (UDR). After over forty-five years of uninterrupted power the DC disintegrated between 1993 and 1998 leaving in its wake a flotilla of small, often unstable and poorly organized parties with no hope of reacquiring the power and influence of the extinct DC. The disintegration of the DC left a vacuum in the Italian political system that was eventually filled through the creation of a party wholly new both in form and content. The history of the Forza Italia party, led by media magnate Silvio Berlusconi, is relatively short. Amid the corruption scandals and proposals for electoral reform Berlusconi began to create his own politically oriented network in In early 1994, during the run-up to the general elections Berlusconi challenged the moderate forces to join together to combat the parties of the left, stating that if they failed to do so he would enter the elections at the head of the new Forza Italia. Immediately following this was the effective dissolution of the DC. Not surprisingly Berlusconi took this as an indication that the centre and moderate parties were incapable of joining together and officially entered the political fray at the head of his new Forza Italia party. In one of the more surprising political success stories in post-war Europe, Berlusconi was able to lead the centre-right to victory in the 1994 elections through an electoral coalition with the regionalist Northern League (LN) in

6 980 Journal of European Public Policy the north and with the right-wing National Alliance (AN) in the south. 7 The 1994 elections marked the complete transformation of the Italian party system. The old key parties of the past, the DC, PCI and even the Socialist Party (Partità Socialista Italiana or PSI) were gone or mere shells of what they had been. Instead new parties (such as the LN and the Forza Italia) or significantly reformed parties (such as the PDS and the AN) were the primary parties. The largest party in the country was wholly new and had been created just months before the election. Unlike the DC and PCI the new parties had almost no social penetration, weak historical ties and in many cases were based on personality more than ideology (Newell 2000; Bufacchi and Burgess 1997). The reform of the electoral system was also a key aspect of this transition. The existing electoral system had been one of the most proportional in existence with over twenty parties holding seats in the lower chamber at times. 8 Following the discovery of the corruption scandals, many of which were linked to campaign finance, there was a movement to implement radical electoral system reform, including the introduction of a majoritarian system with single member districts. A proposal requiring that 75 per cent of the Senate s (and later also the Chamber s) seats be allocated through single member districts and the remaining 25 per cent through proportional representation was overwhelmingly approved by referendum. In June 1993 the results of the referendum were officially adopted by the legislature with some minor changes. The general elections of March 1994 were the first with the new electoral system. The right-wing alliance (above) was the winner, obtaining a majority of seats in the lower chamber and falling just three short of one in the Senate. The centre-left Progressive Alliance did not fare well, due primarily to internal squabbling and the inability to appeal to a broad range of voters. The new electoral laws had a significant impact on the results of the election creating a large distortion between the percentage of votes won and the percentage of seats gained (the right-wing coalition received 42.9 per cent of the vote and 58.1 per cent of the seats). However, the new system failed to live up to expectations and provided neither a significant reduction in the number of parties nor a clear governing majority. In fact, the large parties were actually somewhat less large than in the past. In the five years between 1989 and 1994 the entire Italian political system was transformed. The élite that had governed Italy for nearly half a century were thoroughly delegitimized, many under investigation for corruption or worse. The main political parties of the post-war era had disappeared, either through fragmentation or internal reform and reinvention. In their place was a new kind of party based more on personality than ideology with weak ties to society as a whole. The system of extreme electoral proportionality had been replaced by one based primarily on majoritarian principles. By 1994 a coalition of new and reformed parties of the far right was in power and the old DC and PCI were nowhere to be found. The changes wrought in these five years endured, bringing in a centre-left coalition in 1996 and the return of the Berlusconi coalition in 2001.

7 A. Kreppel: Moving in the right direction? 981 The question is what significance, if any, did these profound changes in the party system of Italy have on the EP, the focal point of partisan activity in the EU? Did the transformation of the Italian party system have any effect on the supranational party system of the EP or the role of the Italian MEPs within it? WHAT KIND OF CHANGE? The impact of domestic political party system change might be expected to manifest itself in two arenas and in two directions as graphically demonstrated in Figure 1. The first direction reflects the influence that Italian MEPs can exert at the supranational level. In this direction the first arena is the EP as a whole, and in particular the EP party system (quadrant A). Here we might expect that changes in the relative balance of power between parties in the domestic arena would affect the EP party system since elections to the EP are organized around national parties. Thus, shifts in electoral fortunes and the creation or dissolution of political parties at the national level potentially impact the relative size of the supranational groups in the EP, which consist of national party delegations. In this context the changes in the Italian party system might have influenced the balance of power between party groups depending on which groups the new parties joined (if any) and which party groups lost members as a result of old Italian parties effectively disappearing. However, because EP elections are independent from national elections these changes can only occur following the former and are less able to directly mirror national change on an annual basis. As a result this line of inquiry is not pursued further here. The second arena in which changes in the domestic sphere may have a significant impact is within the party groups themselves (quadrant B). Since each EP party group is made up of a number of national party delegations there is always some level of internal ideological variation. As the EP has grown and the party groups have assumed members from a broader spectrum of national parties this internal variation has increased. However, EU enlargement is not a prerequisite for increased internal party group diversity. Changes within existing member states can have the same effect if new parties are created that differ from the existing EP party groups but nonetheless seek membership within them. 9 Figure 1 Arenas of potential change

8 982 Journal of European Public Policy Studies of voting in the EP note the generally high level of voting cohesion within party groups (Hix et al. 2003; Hix et al. 2004; Kreppel 2002b). But some also underline the key role of the national party delegations, the tendency of defections from party group positions to be national in character (Kreppel 2002b), and the ability of national parties to influence MEP voting behaviour (Hix 2002). This suggests that changes within the national party system might be reflected at the EP level by a decrease in internal party group cohesion and national delegation defection, or in other words, a decrease in the unity of those party groups affected by the national party system changes. The clearest evidence of increased internal party group dissension would be a reduction in voting cohesion within the EP party groups. Given the frequent requirement of an absolute majority, a reduction in the cohesion of either of the two largest parties could diminish the efficacy of the EP and reduce its ability to bargain with other EU institutions. The impact of changes in the internal unity of party groups with substantial Italian membership can be examined through an analysis of roll-call votes. Of particular interest is a comparison between the internal unity of the European People s Party (EPP) and the Party of European Socialists (PES) before and after Italian party system change, as these two groups were the most directly impacted and are the most powerful groups in the EP. The EPP initially was the group of the DC and later (after initially being refused) became the group of the Forza Europa delegation. Likewise, the PES was influenced by the transformation of the PCI into the PDS, which eventually opted to join the PES where its predecessor had not (remaining in small variously named coalitions of the far left). In both of these cases the addition of the new national Italian party delegation led to an expansion of the ideological spectrum included within the group and potentially to a decrease in party group unity. In addition to these potential changes, there is also the possibility that the influence and role of Italian MEPs underwent a transformation. In particular, it is possible that the instability of the new national parties, the resulting high rate of turnover among Italian MEPs (especially in the 1994 EP elections), and the unclear status of the Forza Italia members led to the marginalization of Italian MEPs within the organizational structure of the EP leading to a decrease in the ability of Italian MEPs to affect decision-making in the EP as a whole (quadrant C). 10 The extent to which this occurred can be measured by examining the representation of Italian MEPs in significant positions within the EP both before and after the shift at the national level looking specifically at the allocation of rapporteurships and committee chairmanships, which are the key prizes that can be broadly distributed among MEPs. A similar situation may have occurred within the party groups themselves and this could also be measured by examining the leadership role of Italian members within the largest groups (EPP and PES) across time (quadrant D). As changes in the role of MEPs within the EP as a whole appear more significant, and to a certain extent reflect the relative power balance within the party groups (as

9 A. Kreppel: Moving in the right direction? 983 their leadership is crucial in the selection process) this is the only aspect of potential change in MEP influence that will be covered here. CHANGES IN MEP BEHAVIOUR AND GROUP UNITY To measure the extent to which the changes at the national level among Italy s parties influenced the internal unity of the EP party groups I analysed voting patterns (during roll-call votes) across the period. 11 In particular, I traced the similarity of voting patterns between Italian MEPs and those of their respective party groups as well as the general agreement between all Italians and the EP as a whole. The focus of the analysis is on those party groups that experienced the greatest changes. This includes the EPP and PES throughout the period, the Communist and left coalition groups during the period through 1992, and the small party groups of the right including the Forza Europa (FE) group during the post-1994 period. 12 The primary goal was to determine the extent to which the voting behaviour of Italian MEPs changed as a result of the party system changes at the national level. Votes were not disaggregated by committee of origin or legislative topic as the primary goal of this research was to first determine if the voting behaviour of the Italian MEPs changed in the aggregate after the transformation of the national party system. 13 Again, given the critical role of the national parties in determining who gets elected to the EP, we should expect that changes in the national parties will be reflected at the EP level. 14 I chart the evolution of voting behaviour among Italian MEPs during six one-year periods. I use 1984 as a base year since this pre-dates any of the changes discussed above. The second year selected is 1988, which marked the beginning of the political earthquake in Italy. The third year is 1990, which immediately followed the decision of the PCI to change its name to PDS and undergo thorough internal reforms. The fourth year selected, 1992, follows on the actual creation of the PDS and its active attempts to join the ranks of European Socialists (rather than Communists). The last two years selected (1995 and 1999) focus instead on the dissolution of the DC and the rise of the Forza Italia after the corruption scandals and electoral reform of the early to mid-1990s. The year 1999 is divided into two groups, one just before the EP elections and one just after. Tables 1 6 report the correlation coefficients for the various years. In each case 100 votes were analysed (for a total of 700 roll-call votes) and the correlation coefficients were determined on the basis of yes and no votes. 15 Since the goal is to compare the behaviour of the Italian members to the other members of their EP party groups and the EP as a whole, I divided the party groups into the Italian members and the rest of the group minus the Italian members; a similar method is used for the EP as a whole. 16 The main aspects to focus on are the correlation coefficients between the Italian members and their EP party groups as well as those with their future party group (i.e. the correspondence between PCI MEPs and those of the PES group or the Forza

10 984 Journal of European Public Policy Table 1 Correlation coefficients measuring unity Ital EPP EPP-Ital Ital PES PES-Ital Ital COM COM-Ital Ital All All-Ital Ital EPP 1 EPP-Ital Ital PES PES-Ital Ital COM COM-Ital ñ Ital All All-Ital Italia members and the EPP group). It is also possible to get a sense of the overall influence of national identity on voting behaviour via the correlation coefficients between Italian members of different party groups. In 1984 (Table 1) the DC and PCI were the undisputed leaders at the national level. The Italian socialists, although still a small party, were increasing in importance within the governing coalition. The Italians were a large component within both the EPP and the Communist groups and a much smaller delegation within the PES. Overall the DC members tended to vote overwhelmingly with the other members of the EPP group (correlation coefficient of 0.911) while the PSI and PCI members voted substantially less often with their groups (0.665 and a striking for the Communists). Interestingly, even at this stage the Italian Communists voted more frequently with the PES than they did with other Communists (correlation coefficient of 0.578). 17 Although change was being discussed nothing formal had occurred by early 1988 (Table 2) and as a result it is not surprising that the patterns of voting cohesion are largely unchanged. There is a reduction in the level of agreement between the Italian members of the EPP and the rest, but the overall level of agreement remains quite high (0.858). The Italian Socialists and Communists also vote more frequently with their EP party groups but still not at the same Table 2 Correlation coefficients measuring unity Ital EPP EPP-Ital Ital PES PES-Ital Ital COM COM-Ital Ital All All-Ital Ital EPP 1 EPP-Ital Ital PES PES-Ital Ital COM COM-Ital ñ0.031 ñ Ital All All-Ital

11 Table 3 Correlation coefficients measuring unity 1990 A. Kreppel: Moving in the right direction? Ital EPP EPP-Ital Ital PES PES-Ital Ital LC LC-Ital Ital All All-Ital Ital EPP 1 EPP-Ital Ital PES PES-Ital Ital LC LC-Ital Ital All All-Ital level as the DC (0.705 and respectively). Similarities in voting between the Italian Communists and the PES increase slightly to (previously 0.578). It is notable that this is much higher than the correlation coefficient of between the PES and the non-italian Communists. Despite the significant changes taking place in the PCI during 1989 and 1990 within the EP very little changed (Table 3). The DC members of the EPP still voted largely with the group as a whole (correlation coefficient of 0.877). The tendency of the PCI/PDS to vote with their new party group increased but this is most likely due to their preponderance within the new group (18/24 members or 75 per cent). The overall tendency to also vote with the PES continued at about the same level (0.630 versus previously) while the tendency of the Italian Socialists to vote with the PES as a whole continued to increase (from to 0.782). By 1992 (Table 4) the transformation of the PCI into the PDS was complete and the corruption scandals were quickly destabilizing the government and the DC and PSI as cohesive parties. However, despite these events, there was no immediate impact on the relative strength of the various Italian delegations or the parties they joined as the composition of the EP remained unchanged since the 1989 EP elections. There was a notable increase in the tendency of the Italian PDS members to vote with the PES group (0.712) rather than their Table 4 Correlation coefficients measuring unity Ital EPP EPP-Ital Ital PES PES-Ital Ital LC LC-Ital Ital All All-Ital Ital EPP 1 EPP-Ital Ital PES PES-Ital Ital LC LC-Ital Ital All All-Ital

12 986 Journal of European Public Policy Table 5 Correlation coefficients measuring unity Ital EPP EPP-Ital Ital PES PES-Ital Ital Right Right-Ital Ital All All-Ital Ital EPP 1 EPP-Ital Ital PES PES-Ital Ital Right* Right-Ital Ital All All-Ital *See note 18. own Left Coalition (LC) (0.551). By and large both the DC and PSI delegations maintain their previous behaviour with a small increase in voting similarity between the Italian PSI and the PES as a whole (from to 0.808). The changes at the national level in 1994 (Table 5) were quite radical. The birth of the so-called Second Republic witnessed the dissolution of the DC and the rise of the Forza Italia; however, even though the 1994 EP election closely followed the transitional national elections the changes in the behaviour of Italian MEPs were less dramatic. On the left the PDS officially joined the PES. The result was a significant increase in the tendency of the PDS members to vote with the PES (from in 1992 when they were in the LC to 0.899). This suggests that, despite the potential increase in the ideological diversity of the PES, the tendency was for PDS members to conform to the will of the group. 18 The EPP members managed to maintain their previously high levels of voting unity with the group, despite the fact that the sixteen Italian members now came from six different national parties. The biggest change was the addition of the strong right-wing Italian contingent and in particular the twenty-six representatives of the new Forza Italia. This group had a very low voting correlation both with other MEPs of the far right and the EPP, which it had aspirations to join (just and respectively). By early 1999 (Table 6a) the Italian party system had basically stabilized around two poles, one centred on the PDS and the other around the Forza Italia. 19 The new political system was largely consolidated including the successful alternation of power between the left and the right. Within the EP, however, there were significant changes in the voting tendencies of the Forza Italia EP members. Although still not formally part of the EPP, the Forza Italia delegation had been actively pursuing membership. There was at the same time a dramatic increase in its tendency to vote with the EPP (0.775) and a still further reduction in its tendency to vote with the other party groups of the far right (0.206). This is similar (albeit more rapid) to what occurred prior to the PDS s integration into the PES six years earlier. The other main Italian

13 A. Kreppel: Moving in the right direction? 987 Table 6a Correlation coefficients measuring unity 1999 (IV Legislature) 1999 Ital EPP EPP-Ital Ital PES PES-Ital Ital Right Right-Ital Ital All All-Ital Ital EPP 1 EPP-Ital Ital PES PES-Ital Ital Right* Right-Ital ñ Ital All All-Ital *See note at foot of Table 6b. Table 6b Correlation coefficients measuring unity 1999 (V Legislature) Ital EPP-ED- Ital Right- FI 1999 (V) EPP-ED Ital Ital PES PES-Ital Right Ital Ital All All-Ital EPP Ital EPP-ED 1 EPP-ED-Ital Ital PES PES-Ital Ital Right* Right-Ital ñ0.092 ñ Ital All All-Ital FI EPP *The Italian right here does not include the Forza Italia members (who joined the EPP) but only the AN and the Lega Nord which are within the Technical Group of Independents (TGI) and the Union of European Nations (UEN). delegations continued to vote cohesively with their groups (with the EPP delegation reaching an all-time high correlation of and the PDS group within the PES dropping slightly to 0.888). The tendency of the Forza Italia MEPs to vote with the EPP increased still further after the 1999 EP elections and their formal incorporation into the newly renamed EPP-ED (Table 6b). Overall the correlation coefficient for the entire Italian delegation fell slightly to from its previous high. More significant, though, is the voting pattern of just the Forza Italia members (consisting of twenty-three of the total thirty-five Italian members of the EPP- ED), which was nearly equal at (compared to the just six months earlier). This was also the pattern for the PDS which once again increased their tendency to vote with the PES (an all-time high of 0.939). Perhaps the most striking aspect of this analysis is how little actually changed in the behaviour of Italian MEPs despite the radical transformations occurring at the national level and the impact that these transformations naturally had

14 988 Journal of European Public Policy on who the Italian MEPs were (personally and ideologically). On the whole the Italian delegations have voted loyally with their EP party groups and this tendency has grown consistently across time. What is most notable is the tendency of the Italian delegations to begin to vote with other party groups when they had a long-term goal of joining that group. This trend is evident over the long run for the PCI/PDS and in a much more condensed way for the Forza Italia delegation. In both cases these delegations strayed from previous voting patterns (in the case of the Forza Italia quite dramatically) to pursue a voting strategy much more in line with the two largest party groups. Furthermore, once they joined these groups formally they continued the trend, achieving an extremely high level of unity with their new party group. It is not hard to understand why the Italian delegations might want to join the two largest groups in the EP. Since direct elections the EPP and the PES have between them controlled between 54 per cent and 66 per cent of the seats. Within the EP benefits and resources such as committee chairmanships and rapporteurships are allocated based on party group size. As a result members are most likely to be able to act effectively and wield real influence if they are within one of the two largest party groups. What is surprising is the extent to which the Italian members seem to have assumed the goals of their new party groups. Given the large size of the EPP and PES the Italian delegations were never able to usurp the party groups decision-making process. In fact, in neither case were the Italians the largest national delegation. So the trend of high voting unity is unlikely to be the result of the Italian delegation having shaped the voting patterns of the group as a whole (more than 150 members from fourteen countries in both cases), but rather represents a real change in the behaviour of the Italians themselves. This is important because it suggests that, contrary to initial expectations, we do not witness national changes impacting the supranational level (beyond the short term). Instead what occurs is, in fact, much more in line with the standard Europeanization literature. The norms and rules at the supranational level appear to effectively impose adaptation and change upon the domestic actors who learn to compromise and work with their European colleagues regardless of what occurs at the national level. The result is that very little changed in the behaviour of the Italian delegations within the EP after an initial transition period. This helps to explain the distinct absence of significant change that we see in terms of the roles of the Italian members within the EP as a whole (quadrant C). CHANGES TO THE ROLES AND INFLUENCE OF ITALIAN MEPs Most activities within the EP are organized on the basis of the party groups (Hix 2002; Kreppel 2002b). Positions of relative influence such as committee chairmanships and rapporteurships are distributed amongst the party groups proportionally using the D Hondt method. In this process each party group is allocated points which it can then spend to buy various positions. Each role

15 A. Kreppel: Moving in the right direction? 989 or task within the EP is assigned a price in points. For example, not all rapporteurships are the same; some may be free requiring no use of points, others may be worth one, two or more points (Corbett et al. 2003; Kreppel 2002b). To investigate the extent to which political change at the national level impacted the influence of the Italian MEPs, I collected information on the allocation of rapporteurships under the co-operation (SYN) and codecision (COD) procedures between 1990 and The EP is able to exert direct legislative influence only under these two procedures so they are clearly the most crucial and represent a growing percentage of the overall workload of the EP. I also examine the allocation of committee chairmanships from 1979 to The data are organized by year for rapporteurships and legislative term for committee chairmanships. Specific information is reported only for the big four countries (which all have an equal number of MEPs) to ensure comparability. Information on rapporteurships is reported in Tables 7a (raw numbers), 7b (percentages) and 8 (distribution by party group) while the data on chairmanship allocations are reported in Tables 9a (raw numbers), 9b (percentages) and 10 (distribution by party group). The results regarding rapporteurships are unclear to say the least (Tables 7a and 7b). There does not appear to be any correlation between events at the national level, or changes in party group membership at the EP level and the yearly distribution of rapporteurships. Italians received their highest number of rapporteurships in 1995, 1996 and 2000, but these years are not particularly significant otherwise. 22 There is no noticeable impact of the formal accession of the PDS to the PES group in 1993, nor does the formal adhesion of Forza Italia members to the EPP group in 1999 appear to have much impact, although this may account for the rise in 2000 when the Italian EPP members received four rapporteurships (Table 8). As is to be expected members of the two largest groups receive the lion s share of Italy s rapporteurships (twentyone out of thirty-five or 60 per cent). The greater number allocated to members of the EPP, especially after 1994 (and the break-up of the DC), is surprising but, on its own, not a substantial enough finding to allow us to draw any general conclusions. Overall, what is perhaps most notable is the very low number of rapporteurships allocated to Italians regardless of the year (never more than 15.2 per cent). Other countries such as Germany and the UK average 19 per cent and 17 per cent respectively, while Italy averages just 7 per cent (France 8 per cent). On the whole Italian (and French) members also suffer from low turnout rates while German and UK members tend to have the highest attendance rates and this may be a critical factor when deciding who to give a rapporteurship to (Kreppel 2002b; Hix et al. 2004). On the other hand, Italians do much better in winning committee chairmanships (Table 8). In fact, across all five legislative terms Italian members have secured as many or more committee chairmanships as members from any other country. However, there is very little change over time that can be tied

16 Table 7a Rapporteurship distribution among the big four (SYN and COD) Total France Germany Italy UK Other Total

17 Table 7b Rapporteurship distribution among the big four in % (SYN and COD) Average France 11.1% 6.3% 10.5% 10.0% 13.0% 8.1% 13.0% 9.6% 4.6% 5.9% 0.0% 5.7% 8% Germany 0.0% 18.8% 26.3% 13.3% 21.7% 22.6% 28.3% 30.8% 20.0% 17.6% 21.6% 11.4% 19% Italy 11.1% 0.0% 5.3% 0.0% 6.5% 11.3% 15.2% 7.7% 3.1% 2.9% 9.5% 5.7% 7% UK 0.0% 50.0% 10.5% 16.7% 23.9% 16.1% 8.7% 15.4% 21.5% 8.8% 17.6% 14.3% 17% Other 77.8% 25.0% 47.4% 60.0% 34.8% 41.9% 34.8% 36.5% 50.8% 64.7% 51.4% 62.9% 49%

18 Table 8 Party group allocation among Italian rapporteurs (SYN and COD) Total GUE/NGL EPP PES ELDR FE UPE TDI Other Total

19 Table 9a Committee chairmanships * A. Kreppel: Moving in the right direction? 993 I II III IV V Total France Germany Italy UK Other Total *Each two and a half year term is counted individually since this represents a reappointment. Table 9b Committee chairmanships (%)* I II III IV V Average France 17.1% 13.2% 11.4% 11.6% 8.6% 12.4% Germany 22.9% 18.4% 13.6% 14.0% 20.0% 17.8% Italy 25.7% 21.1% 15.9% 18.6% 20.0% 20.3% UK 17.1% 18.4% 15.9% 14.0% 17.1% 16.5% Other 17.1% 28.9% 43.2% 41.9% 34.3% 33.1% *Each two and a half year term is counted individually since this represents a reappointment. to national events. Although there is a decline in the overall percentage of chairmanships obtained from the first to the third legislature (Table 9b) this has more to do with EU enlargement and the need to distribute the positions among a larger number of MEPs overall than any effects from domestic political change. 23 In fact, similar patterns of decrease are evident among the French, German and UK delegations as well. The distribution of Italian committee chairmanships among the party groups is also remarkably consistent with the only change being a distinct reduction in the number of chairmanship positions going to the Italian members of the Communist/Left coalition group in the run-up to the PDS joining the PES (Table 10). However, there is no subsequent increase in the number of chairs allocated to Italian members of the PES after 1993 (when the PDS joined). There does not appear to be any clear impact of the rise of the Forza Italia or the functional end of the DC. Although the Forza Europa group within the EP was never allocated a chairmanship on its own the Italian EPP members (be they DC, post-dc parties or Forza Italia members) received four chairmanship posts per legislature from 1989 on. Thus, once again we are left with the fact that the dramatic changes that occurred at the national level and the wholesale restructuring of the Italian

20 994 Journal of European Public Policy Table 10 Party group allocation among Italian chairs* I II III IV V Total COM/LC EPP PES ELDR Total *Each two and a half year term is counted individually since this represents a reappointment. party system seem to have had little or no effect on the role of Italian MEPs within the EP itself. Both before and after these changes at the national level Italian MEPs received comparatively few rapporteurships and the ones they did receive were largely allocated to members of the EPP and PES. Although on the whole Italian members were much more successful in obtaining committee chairmanships (relative to other national representatives) there was once again no pattern of change over time that can be clearly linked to the political transformations experienced at the national level. 24 The question of course is why? And what does this singular lack of impact, combined with the similar absence of significant change in MEP voting behaviour, mean both for the Europeanization literature and for the future of the EU more generally? CONCLUSIONS: IS IT A ONE-WAY STREET? Despite the truly revolutionary reforms in the Italian political party system that took place in the 1990s very little changed either in the behaviour of Italian MEPs (in terms of voting) or in their relative access to positions of influence within the EP. Across the period analysed Italian MEPs tended to vote with their party groups and this tendency increased over time with the sole exception of the Forza Italia members right after the 1994 elections. When Italian national party delegations aspired to join one of the two large party groups they began to vote like them and continued to vote with them at high levels after formally joining. Historically Italians have received below their fair share of rapporteurships, but more than their fair share of committee chairmanships. The changes in the domestic political sphere appear to have had little or no influence on these overall trends. Nor have they impacted the distribution of these benefits significantly beyond a decrease in the numbers of positions given to the far left. Overall the changes that have occurred have been either temporary, minimal or both. This is somewhat surprising given the depth of the change that occurred at the national level and the critical role played by the party groups in the EP. There are some possible explanations, which are important in light of both future enlargements and the general debate about Europeanization. The first

21 A. Kreppel: Moving in the right direction? 995 is simply that no matter how significant the changes were at the national level they would not significantly affect the EP because in the end the 14 per cent of MEPs made up by the Italian delegation is simply not significant enough to instigate broad change. It is likely that had these reforms taken place back in the 1970s, before enlargement when the Italian delegation accounted for a much larger percentage of the whole EP, the implications (especially for the EPP) would have been more significant. This does suggest that as the EU continues to enlarge and each individual country makes up a smaller percentage of the overall membership of the EP, changes at the national level will become increasingly less important. While this explanation may be partially accurate, it is unsatisfactory in explaining why the Italian MEPs themselves did not change their behaviour, especially following the 1994 or 1999 elections. 25 A more interesting interpretation of the results of this analysis is based not on size, but rather on the impact of institutionalization and norms. The EP is a well-developed legislature with an extensive network of formal and informal norms and rules. Some, like the broad application of the D Hondt method (between and within the party groups), limit the ability of actors within the EP to react strongly to changes within national delegations. Others, like the emphasis on consensus building and broad pro-integrationist coalitions, counteract strong variations in ideology that may be much more significant at the national level. In fact, within the EP there is an extremely high level of overall cross-party group consensus (see the correlation coefficients between Ital All and EPP-Ital in Tables 1 6) and the EPP and the PES also tend to vote together more often than not (Kreppel 2000; Kreppel and Hix 2003; Hix et al. 2004). This norm of compromise and moderation may have many roots (including the need to present moderate proposals able to win support from the other EU institutions), but the important aspect for the purposes of this research is the extent to which new actors within the EP succumb to it. The evidence presented here suggests that both the PDS and Forza Italia members changed their behaviour to work with and then within the two largest groups rather than acting independently (which would have left them largely isolated and ineffective) or changing the actions of others (which most likely would have been impossible given the numbers and diversity of the other actors involved). If this is what occurred, it presents a picture more or less in line with standard Europeanization theories. Instead of the national actors affecting change at the supranational level, they were in fact Europeanized, inculcated into the norms and controlled by the rules that govern activity at the supranational level within the EP. This suggests a much higher level of resiliency among the EU institutions than might have been expected given their relative youth and the constant level of change and reform they have experienced over the past fifty years. It also presents an optimistic alternative for many of the negative scenarios associated with the enlargement process. Although the expansion to include ten new member states in May 2004 will be far more traumatic than the changes wrought by the political upheavals